Partial Hepatectomy Induced Long Noncoding ... - Semantic Scholar

RESEARCH ARTICLE

Partial Hepatectomy Induced Long Noncoding RNA Inhibits Hepatocyte Proliferation during Liver Regeneration Lulu Huang, Sagar S. Damle, Sheri Booten, Priyam Singh, Mahyar Sabripour, Jeff Hsu, Minji Jo, Melanie Katz, Andy Watt, Christopher E. Hart, Susan M. Freier, Brett P. Monia, Shuling Guo* Department of Antisense Drug Discovery, Isis Pharmaceuticals Inc., Carlsbad, CA, 92010, United States of America * [email protected]

Abstract

OPEN ACCESS Citation: Huang L, Damle SS, Booten S, Singh P, Sabripour M, Hsu J, et al. (2015) Partial Hepatectomy Induced Long Noncoding RNA Inhibits Hepatocyte Proliferation during Liver Regeneration. PLoS ONE 10(7): e0132798. doi:10.1371/journal.pone.0132798 Editor: Zhuang Zuo, UT MD Anderson Cancer Center, UNITED STATES Received: May 2, 2015 Accepted: June 19, 2015 Published: July 24, 2015 Copyright: © 2015 Huang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Data were deposited in the Gene Expression Omnibus (Accession no. GSE70343). Funding: The work was supported by Isis Pharmaceuticals in the form of salaries for authors LH, SSD, SB, PS, MS, JH, MJ, MK, ATW, CEH, SMF, BPM and SG, but the company did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.

Liver regeneration after partial hepatectomy (PHx) is a complex and well-orchestrated biological process in which synchronized cell proliferation is induced in response to the loss of liver mass. To define long noncoding RNAs (lncRNAs) that participate in the regulation of liver regeneration, we performed microarray analysis and identified more than 400 lncRNAs exhibiting significantly altered expression. Of these, one lncRNA, LncPHx2 (Long noncoding RNA induced by PHx 2), was highly upregulated during liver regeneration. Depletion of LncPHx2 during liver regeneration using antisense oligonucleotides led to a transient increase in hepatocyte proliferation and more rapid liver regeneration. Gene expression analysis showed that LncPHx2 depletion resulted in upregulation of mRNAs encoding proteins known to promote cell proliferation, including MCM components, DNA polymerases, histone proteins, and transcription factors. LncPHx2 interacts with the mRNAs of MCM components, making it a candidate to regulate the expression of MCMs and other genes posttranscriptionally. Collectively, our data demonstrate that LncPHx2 is a key lncRNA that participates in a negative feedback loop modulating hepatocyte proliferation through RNARNA interactions.

Introduction Although long noncoding RNAs (lncRNAs) such as H19 and Xist were first discovered decades ago, only in the last few years have the advances in transcriptome sequencing technologies led to the discovery of thousands of previously unannotated lncRNAs [1, 2]. A recent update by the GENCODE consortium annotated 9,277 lncRNA genes that result in 14,880 transcripts [3]. Currently, lncRNAs are defined by two features: 1) the transcript is longer than 200 nucleotides and 2) does not appear to have coding potential. The first criterion of length is arbitrary and based on the purification method. Essentially, these RNAs are defined by what they are not rather than their function [2]. LncRNAs are often expressed in a tissue- and developmental

PLOS ONE | DOI:10.1371/journal.pone.0132798 July 24, 2015

1 / 19

LncRNA in Liver Regeneration

Competing Interests: The authors LH, SSD, SB, PS, MS, JH, MJ, MK, ATW, CEH, SMF, BPM and SG, are employees of Isis Pharmaceuticals, the funder of this study. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.

stage-specific manner [1, 4], and levels of many are altered in disease states [5, 6]. Some lncRNAs have been shown to be essential for life and development, such as Xist [7], Fendrr [8, 9], and Braveheart [10]; others play important roles during disease progression [5], such as Malat1 [11–14] and Pvt1 [15]. Initially, it was thought that the main function of lncRNAs is to regulate transcription through influencing the epigenetic status of the chromatin [16]. More recently, however, lncRNAs that regulate almost every step of gene expression, including RNA processing, RNA stability, and translation, have been identified [2]. Yet the functions of most lncRNAs are unknown. Previous studies have implicated lncRNAs in the regulation of cell proliferation [17]. To further understand the roles of lncRNAs in cell proliferation, we analysed lncRNA expression in a mouse two-thirds partial hepatectomy (PHx) liver regeneration model [18]. Liver regeneration after PHx is a very complex and well-controlled process, and requires participation of all mature liver cell types with hepatocytes being the main players [19–23]. Immediately following surgery, growth factors and cytokines work together to induce mature hepatocytes to re-enter cell cycle, which in turn triggers cell proliferation of the other cell types in the liver. Within 72 hours, hepatocytes complete 1 to 2 rounds of synchronized proliferation, and liver mass and function is fully restored in approximately 10 days. Liver mass is precisely controlled, as there is no over growth of the liver in response to PHx. A cascade of robust transcription regulation triggered by cytokine and growth factor signalling regulates this well orchestrated biological process [22, 24]. We performed genome-wide gene expression profiling to identify lncRNA expression changes during liver regeneration after PHx. We found that approximately 400 lncRNAs were differentially expressed after PHx. Interestingly, one lncRNA, LncPHx2, whose expression is induced after PHx, was shown to negatively regulate hepatocyte proliferation through inhibition of the genes that promote cell growth.

Materials and Methods Animal experiments PHx was performed as described before [18]. In brief, male Balb/c mice (Charles River Laboratories), 7~9 weeks of age were under isofluorane anesthesia (2% in air restrainer for induction and 1–2% via nose cone for maintenance). Left literal lobe and median lobe of the liver were removed with two separate ligatures. For experiments involving antisense oligonucleotide (ASO) treatment, mice were injected subcutaneously with LncPHx2_ASOs, control ASO, or PBS as indicated in the main text. The DEN-induced mouse HCC model was previously described [25]. In brief, male C57BL/6 mice, 15 days of age, were injected intraperitoneal with 25 mg/kg diethylnitrosamine (DEN, Sigma). A pool of DEN-injected BL/6 mice was maintained for 8 months to allow tumor formation, and then treated subcutaneously with ASOs or control reagents for 3 months before sacrificing and data collection. Animals were euthanized by exsanguination under Isoflurane inhalation followed by cervical dislocation. All animal husbandry and procedures were approved by the Institutional Animal Care and Use Committee at Isis Pharmaceuticals.

Microarrays, RNA sequencing, and data analysis Genome-wide profiling of mRNA and lncRNA expression changes during liver regeneration were performed using the NCode Mouse Non-coding RNA Microarray (Invitrogen). Data were normalized for intensity dependent variance using the vsn package (Bioconductor). Differentially expressed genes were identified and clustered using the maSigPro (microarray Significant Profiles) R-package. A two-step regression were performed to first identify significantly differentially expressed genes (FDR = 0.05), and then to identify the conditions

PLOS ONE | DOI:10.1371/journal.pone.0132798 July 24, 2015

2 / 19

LncRNA in Liver Regeneration

that show statistically significant differences (alfa = 0.05, regression step = two.ways.backward). Genes were further filtered by goodness of fit gene profiles against gene regression models (Rsquared > = 0.6), then aggregated into 9 clusters using hclust [26, 27]. To identify LncPHx2regulated genes, mouse liver RNAs were analysed using Illumina True-seq protocol. Reads were processed using STAR [28]. Differential gene expression calls were made using cuffdiff [29]. GSEA analysis was done with pre-ranked gene list by expression [30].

Antisense oligonucleotides Antisense oligonucleotides used in this study were chemically modified with phosphorothioate in the backbone and constrained ethyl (cET) modifications in the wings with a central 10-nucleotide deoxy gap (3-10-3 gapmer). Oligonucleotides were synthesized using an Applied Biosystems 380B automated DNA synthesizer (PerkinElmer Life and Analytical Sciences, Applied Biosystems) and purified as previously described [31, 32]. ASO sequences are as follow: control ASO, 3-10-3 cET gapmer, 5’- GGCTACTACGCCGTCA-3’; LncPHx2_ASO1, 310-3 cET gapmer, 5’-AACTTCAAGTAACAGG-3’; LncPHx2_ASO2, 3-10-3 cET gapmer, 5’-AGGCATAACTTCAAGT-3’.

Cell culture and transfection Mouse cell lines MHT [33], BNL.CL2 (ATCC) and Hepa1-6 (ATCC) were cultured in DMEM containing 1% L-glutamine, 10% fetal bovine serum, and 100 units/ml penicillin/streptomycin in 5% CO2 at 37°C. ASOs were added to the culture media 5–12 hours after seeding cells at the indicated concentrations. Cells were harvest 48 hours after ASO-treatment.

Plasma chemistry analysis Blood samples were collected by cardiac puncture at time of sacrifice. Plasma chemistry values were measured on the AU480 Clinical Chemistry Analyzer (Beckman Coulter).

RNA analysis Cultured cells were lysed and the total RNA was extracted with Qiagen RNeasy columns. Animal tissues were homogenized in guanidine isothiocyanate solution (Invitrogen) supplemented with 8% 2-mercaptoethanol (Sigma-Aldrich). Total RNA was prepared using the Qiagen RNeasy columns. Quantitative real-time PCR (qRT-PCR) was performed using an ABI stepone sequence detector. Primer probe sequences are as follow: LncPHx1, forward, 5’TGGATTTGGAAGCTTTGAGTGA-3’, reverse, 5’-CGTCTTTTCTCGGTGCTTGAT-3’, probe, 5’-FAM-CAGACACATGTTCCTCTTCCTCCTGCTCAX-TAMRA-3’; LncPHx2, forward, 5’TGTTGCAGTGTGGTCCAGAGA-3’, reverse, 5’- CTGCTTCTTCTTCAGCAATGGAT-3’, probe, 5’-FAM- AGCCAGCCTTTTTGCTGTGGATCCCX-TAMRA-3’; LncPHx3, forward, 5’- GCACAGCACACTCAGAATTACAAA-3’, reverse, 5’- CCGCCTTTAATCCTAGCACTTG3’, probe, 5’-FAM- ATGTATCCCTGGCTGGCTTGTAACCCAX-TAMRA-3’; LncPHx4, forward,5’- ACGCACCTTCCCCTGTCTT-3’, reverse 5’- TCCGCCTTCTCCATTTTGTG-3’, probe 5’-FAM- TTTGCCCTGTGTCCTTCTGTCTCCTGTT-TAMRA-3’; LncPHx5, 5’GGGCTCCTCATGTGTTCG-3’, reverse, 5’-GGAATGGCAGAACTTCAGGA-3’, probe, 5'FAM- TGGAAGGC-BHQ-1-3’; LncPHx6, forward, 5’-TGCCTTTGGCATTCTTTGTATCT3’, reverse, 5’- GCAGTGCTGGTCCTCTGTGA-3’, probe, 5'-FAM-CTGCGTTTCACAG CAGCAGCCATCTAG-IOWA-BLACK (w/ internal ZEN) -3’, Ccnyl1, forward, 5’TCGCTCCTTAGCAGATGACAAC-3’, reverse, 5’-CTTGAAATGGCCTCTAGGTTCTGT-3’, probe, 5'-FAM- ACCTGAATTTTCTGTTTGCTCCTCTCAGCA-IOWA-BLACK (w/

PLOS ONE | DOI:10.1371/journal.pone.0132798 July 24, 2015

3 / 19

LncRNA in Liver Regeneration

internal ZEN) -3’, Gapdh, forward, 5’-GGCAAATTCAACGGCACAGT-3’, reverse, 5’GGGTCTCGCTCCTGGAAGAT-3’, probe, 5’-FAM- AAGGCCGAGAATGGGAAGCTTGT CATCX-TAMRA-3’. Taqman assay for Ccne1 (Mm00432367_m1), Mcm3 (Mm00801872_m1), Aurkb (Mm01718146_g1), Ccnb1 (Mm03053893_gH), and Rbl1 (Mm01250721_m1) were purchased from life technologies.

Histological analysis Animal tissues were collected, fixed in 10% buffered formalin, and paraffin embedded. Immunohistochemical analyses were performed using a Ki67-specific antibody (Thermo RM9106-S) following manufacturer’s protocols. BrdU (Life Technologies, 00–0103) labelling was detected using BrdU in situ detection kit (BD, 550803). Single molecule RNA in situ hybridization was done using QuantiGene ViewRNA Assays (Affymetrix) following manufacturer's instructions.

RNA interactome analysis RNA interactome analysis was done as previously described [34]. In brief, fifteen biotinylated antisense DNA probes targeting the LncPHx2 RNA were designed using the online designer at http://www.singlemoleculefish.com. Probes were divided based on their postions into two pools, odd pool (odd numbered probes) and even pool (even numbered probes). Eight biotinylated antisense DNA probes targeting the LacZ mRNA were used as negative control. Cells were fixed with 1% glutaraldehyde, and lysed in lysis buffer (50 mM Tris, pH 7.0, 10 mM EDTA, 1% SDS, added just before use: dithithreitol (DTT), phenylmethylsulphonyl fluoride (PMSF), protease inhibitor and Superase-In). Lysates were sonicated using Bioruptor (Diagenode) until completely solubilized. Cell lysate were diluted three times using hybridization buffer (500 mM NaCl, 1% SDS, 100 mM Tris, pH 7.0, 10 mM EDTA, 15% formamide, added just before use: DTT, PMSF, protease inhibitor, and Superase-In). Pooled probes (100 pmol) were added (1ul to 1ml of undiluted lysate), and lysate were incubated end-to-end rotation at 37°C overnight. Biotinlated probes and its associated RNPs were then captured using streptavidin-magnetic C1 beads (100ul beads/100pmol probes) and magnets (invitrogen) after wash (2×SSC, 0.5% SDS, fresh PMSF added, 5 mins at 37°C, repeat 5 times). For RNA elution, beads were treated first using proteinase K (1mg/ml, Ambion. proteinase K buffer:100 mM NaCl, 10 mM Tris, pH 7.0, 1 mM EDTA, 0.5% SDS. 50°C, 45 min, followed by boiling for 10 min). RNA was then isolated using Trizol reagent and was subject to DNase treatment (TURBO DNase kit, Ambion). qPCR analysis of LncPHx2 and Gapdh levels were performed to determine the efficiency and purity of the experiment. cDNA libraries for RNA-seq were made using SMARTer Universal Low Input RNA Kit (Clonetech). RNA sequencing was done on an Illumina Hiseq sequencer (single-end, 100-bp read length, 8 million reads). Sequences of probes targeting LncPHx2 are as follow: #1, 5’-ATGGAAACCAGAATTCGCGC-3’; #2, 5’-CGAGTAA CAAACTGCCGCAG-3’; #3, 5’-AAAACCAACTCTTCACCAGG-3’; #4, 5’-CATGGAGAGAC CAAACTGCT-3’; #5, 5’- TGAGCAAAGGGAAGCTGTCA-3’; #6, 5’- ACATAGTTCTAGCA GATGCT-3’; #7, 5’- AGGCCAGAAAATGTCCAGAC-3’; #8, 5’- AACACATCCCTTTAT CTTCT-3’; #9, 5’- ATCCACAGCAAAAAGGCTGG-3’; #10, 5’-TCTTCAGCAATGGAT GGTGA-3’; #11, 5’-TAGTGTCAGGTGTGTTTGAC-3’; #12, 5’-GTGGAGAAGGGTGAGAA GAC-3’; #13, 5’-TAGGTATTTTTCAGTTCTGT-3’; #14, 5’- AATGCTAAAAGCAGGGGA TC-3’; #15, 5’-AGTTTAGAGAAGTATGCCAT-3’. Sequences of control probes targeting LacZ are as follow: #1, 5’-ATTAAGTTGGGTAACGCCAG-3’; #2, 5’-AATAATTCGCGTCT GGCCTT-3’; #3, 5’- ATCTTCCAGATAACTGCCGT-3’; #4, 5’-AACTGTTACCCGTAGGT AGT-3’; #5, 5’- ACCATTTTCAATCCGCACCT-3’; #6, 5’- TGGTTCGGATAATGCGAACA-

PLOS ONE | DOI:10.1371/journal.pone.0132798 July 24, 2015

4 / 19

LncRNA in Liver Regeneration

3’; #7, 5’- ATTTGATCCAGCGATACAGC-3’. Enrichment peaks were identified as described [34] with the following modifications. Coverage was computed at each position in the genome and an enrichment score (EScore) was calculated by computing the minimum coverage, scaled by number of mapped reads, between independently selected pull-down probe sets. Instead of an input sample, the minimum coverage score was normalized to lacZ probe set coverage to select against non-specific pull-down peaks. Adjacent EScores were merged using bedtools, and a mean EScore was determined. The average log2(EScore) and standard deviation of log2 (EScore) were 0.285 and 0.410, respectively. These regions had a minimum log2 (EScore) of 2, which corresponded to an enrichment greater than 4 standard deviations above the mean. The top 500 EScore regions were used in a de novo motif search in MEME [35]. To ensure that input peak sequences were not limited by narrow peak boundaries, the minimum input length was fixed at 100 bp (i.e., 50 bp flanking the peak centre). Reverse motif search using MAST was performed against the gene transcripts differentially expressed upon treatment with LncPHx2_ASO1, using unchanged gene transcripts as control [36].

Statistics Tow-tailed independent Student’s t test was used for statistical analysis in this study.

Results Genome-wide lncRNA expression profiling during mouse liver regeneration after 2/3 PHx To identify lncRNAs that regulate cell proliferation during liver regeneration, we analysed lncRNA expression profiles in mouse liver tissue collected at 4, 12, 36, and 72 hours after PHx as synchronized hepatocyte proliferation occurs during this time window [37]. Analysis using the NCode Mouse Non-coding RNA Microarray (Invitrogen) revealed that 3653 mRNAs and 465 putative lncRNAs were differentially expressed compared to transcripts isolated from livers of control sham operated mice. Differentially expressed genes were categorized into nine clusters based on their expression patterns after PHx (Fig 1A and S1 Table). Clusters 1, 4, 8, and 9 contain genes that were significantly upregulated after PHx. Clusters 2, 3, and 6 contain genes that were significantly downregulated. Clusters 5 and 7 contain genes that were down- (cluster 5) or up- (cluster 7) regulated immediately after PHx and then the reverse at later time points. KEGG pathway analysis was performed on the coding genes in each cluster. We found that clusters 1and 4 were significantly enriched in genes regulating cell-cycle and cell proliferation (Fig 1B). We hypothesized that the lncRNAs within these two clusters might also affect cellcycle and cell proliferation and, therefore, focused on the lncRNAs within these two clusters. We manually curated putative lncRNA transcripts in clusters 1and 4 by using the UCSC genome browser to look for supporting evidence of mouse expressed sequence tags (ESTs) [38]. We also evaluated the coding potential of these putative lncRNAs using PhyloCSF [39]. qPCR primer/probe sets were designed to amplify the ten most upregulated transcripts that appeared to be lncRNAs based on supporting ESTs and lack of coding potential. Six of the ten lncRNAs were confirmed to be significantly upregulated during liver regeneration (Fig 1C). We named these 6 lncRNAs lncRNA induced by PHx 1–6 (LncPHx1-6) (Fig 1C). Analysis of data obtained on nine mouse tissues available in the Encode RNA-seq database [40] showed that most of these lncRNAs are expressed in tissue-specific manners, with low to median expression in normal mouse liver (Fig 1D).

PLOS ONE | DOI:10.1371/journal.pone.0132798 July 24, 2015

5 / 19

LncRNA in Liver Regeneration

Fig 1. Gene expression profiling of mouse liver regeneration after PHx. (A) Differentially expressed mRNAs and putative lncRNAs were clustered based on their expression pattern during liver regeneration after PHx. Normalized average probe intensity was plotted over the time course of liver regeneration. Cluster 1 contains 401 mRNA and 30 lncRNA transcripts. Cluster 2 contains 471 mRNA and 91 lncRNA transcripts. Cluster 3 contains 610 mRNA and 110 lncRNA transcripts. Cluster 4 contains 385 mRNA and 46 lncRNA transcripts. Cluster 5 contains 146 mRNA and 17 lncRNA transcripts. Cluster 6 contains 410 mRNA and 73 lncRNA transcripts. Cluster 7 contains 254 mRNA and 37 lncRNA transcripts. Cluster 8 contains 330 mRNA and 17 lncRNA transcripts. Cluster 9 contains 646 mRNA and 44 lncRNA transcripts. Sham: liver RNAs of mice subjected to sham surgery. PHx: liver RNAs of mice subjected to PHx surgery. n = 5 for each time point. (B) Pathway analysis of the eight gene clusters using David KEGG pathway tools. Cluster 5 has no enriched pathway (not

PLOS ONE | DOI:10.1371/journal.pone.0132798 July 24, 2015

6 / 19

LncRNA in Liver Regeneration

shown). * Pathway with FDR